mysingingmonstersfandomcom-20200222-history
User blog:BunsenH/Entbrat Stats Phase 1, Report 1
As you may have noticed, my "minions" and I have been working on the statistics of Ghazt breeding for some time. But there are lots of other questions of the "so what are the chances of that?" kind. A few months ago, I asked if the "minions" were willing to do a bit more work for the sake of science (and for the sake of doing more than just the same thing over and over), and they were up for it. Also, doing a different breeding experiment will eventually let us look at the effects of Wishing Torches — are they always the same, or do they depend on what you're trying for? So: What are the odds of successfully breeding an Entbrat? By any standard, they're much easier to breed than Ghazts are. This much is obvious without taking the trouble to record data; it usually takes someone only a few tries to get an Entbrat. ... Unless they try breeding one from two 2-element monsters. That doesn't seem to work well, according to the Entbrat article. People who read that article probably don't even try the 2 + 2 approach, or if they do, they give it up quickly. So we were wondering just what the odds are. Presumably there's a reason for that, from the game-design point of view. After a breeding attempt, one has to wait for the egg to come out of the breeder before trying again. (And wait for the previous egg to hatch in the nursery, before the new egg can move along.) If one breeds a T-Rox with a Potbelly, the delay will be 8 hours if a T-Rox comes out, 2 hours for a Potbelly, and of course 24 hours if the breeding is successful. For, say, a Furcorn and Fwog, the delay from a failure will be 8 hours for a Furcorn, 30 minutes for a Fwog. In other words, the average delay before one can try again is a lot shorter for that 2 + 2 combination. So it makes sense, in a "keep things fair" way (or a "don't give the players a short cut" way), for the odds of success to be lower for the 2 + 2. Okay, turn that around. If one breeds a Bowgart with a Noggin, an unsuccessful breeding attempt gives a delay of 12 hours for the Bowgart, but only 15 seconds for the Noggin. You can just keep trying, over and over, until you get either a Bowgart or an Entbrat working its way through the breeder. So what's the "fair" / "balanced" success probability there? Also keeping in mind that there can be a pile-up at the Nursery, of course. So the "minions" were split up into three groups. One group is working on the T-Rox + Potbelly combination, one group is doing Furcorn + Fwog, and one group is doing Bowgart + Noggin. All of the monsters used for breeding are Level 10, and 25% happy. (So far as we were able to tell in the Ghazt breeding, neither monster level nor monster happiness made any difference, but we want to be as consistent as possible.) The breeding monsters are right beside each other... again, just to be consistent. As you can see, groups 1 and 3 have just about exactly the same chance of success, and breakdown of outcomes. The breeder spits out the parent with the shortest breeding time about 40% of the time, the other parent about 25%, and the Entbrat takes the rest. I was surprised to see just how poor the chance of success was for group 2! It's only about 7%, lower than the others by a factor of about 4.5 . And here, it's the slower parent, the Furcorn, that has the highest chance of coming out, roughly 50%. So: if you're trying to breed an Entbrat, you should definitely stay away from the Fwog + Furcorn combination; the other two we tried are much better. We don't know yet if we can generalize to "avoid all 2-element + 2-element combinations; all 3-element + 1-element combinations are about the same". We'll be looking into that later, after we've refined our statistics for the current combinations. A brief note about our statistics: All of the work we've done about Ghazts had its uncertainty calculations based on the assumption that there was either success (a Ghazt coming out) or failure (always a T-Rox coming out). Here there are three possible outcomes: one of the two parents, or an Entbrat. We can extend the logic to this case. The uncertainty for, say, the fraction of Furcorns is based on "success" being a Furcorn result (the x'' in that derivation having a value of 1) and "failure" being some other result (that ''x having a value of 0). The uncertainties in our results are dropping relatively quickly, thanks to the relatively rapid breeding cycles and moderately high chances of Entbrat breeding (compared to the Ghazt's 1%). We do want to keep on for a while longer, so when we start adding Wishing Torches, we'll be able to tell the difference if the torch's effect is always only a ¾% increase to the chance of success. As always, my thanks to the "minions" for their work. A big project becomes much easier when it's distributed among lots of people. Category:Blog posts